EP0218230A2

EP0218230A2 - Method for producing mucoid and phage resistant group N streptococcus strains from non-mucoid and phage sensitive parent strains

Info

Publication number: EP0218230A2
Application number: EP86113843A
Authority: EP
Inventors: Ebenezer R. Vedamuthu
Original assignee: Microlife Technics Inc
Current assignee: Microlife Technics Inc
Priority date: 1985-10-11
Filing date: 1986-10-06
Publication date: 1987-04-15
Also published as: US5066588A; NZ217722A; US4918014A; US5672477A; EP0218230A3; DE218230T1; ES2002621A6; AU6119586A; CA1311202C

Abstract

A method for imparting phage resistance to phage sensitive strains of Streptococcus group N is described. The method involves transferring plasmid encoding for production of a mucoid substance (Muc⁺) into the phage sensitive strain. Even if the Muc⁺ plasmid is removed by curing at elevated temperatures the strains remain resistant to phage. The resulting resistant strains are novel amd are used for fermentations, particularly milk fermentations.

Description

BACKGROUND OF THE INVENTION

(l) Field of the Invention

The present invention relates to a method for producing phage resistant bacteria from phage sensitive strains of Streptocococcus in N group of this genus. Further the present invention relates to novel bacterial compositions including phage resistant strains of Streptocococcus in N group of this genus derived from phage sensitive strains.

(2) Prior Art

The occurrence of lactic streptococci that produce a mucoid, ropy texture in milk is well documented (Hammer, B.W., Iowa Agr. Expt. Sta. Research Bul. 74:260-270 (l923)). Such ropy lactic streptococci are used in Scandinavian fermented milks called taette (Foster, E.M., et al., Dairy Microbiology, p. l4-l5, 48 and 332, (l957); Rasic, J. L. et al., Yoghurt-Scientific grounds, technology, manufacture and preparations, p. l94 (l978)), Swedish lang mjolk (Bottazzi, V., Biotechnology, Vol. 5, p. 328, 345-346 (l983); Macura, D., et al., J. Dairy Sci. 67:735-744 (l984)) and Finnish villii (Saxelin, M., et al., Canadian J. Microbiol. 25:ll82-ll87 (l979)). Forsen, R., Finnish J. Dairy Sci. 26:l (l966) isolated mucoid strains of all three lactic streptococci, namely, Streptococcus cremoris, Streptococcus lactis and Streptococcus lactis subsp. diacetylactis, from Finnish villii.
The instability of mucoid characteristic in lactic streptococci has been observed by several investigators (Foster, E. M., et al., Dairy Microbiology. p. l4-l5, 48 and 332 (l957); Hammer, 8. W., Iowa Agr. Expt. Sta. Research Bul. 74:260-270 (l923); and Macura, D., et al. J. Dairy Sci. 67:735-744 (l984)). Foster et al reported that mucoid lactic streptococci gained or lost the slime-producing property "capriciously". Macura and Townsley found that ropy lactic streptococci lost the mucoid property after l0 or l2 serial transfers; some strains became non-mucoid even after six transfers. Brooker (Brooker, B.E., J. Dairy Research 43:283-290 (l976)) working with a pure milk culture of a ropy S. cremoris strain observed considerable variations in the proportion of cells producing extracellular capsular material. Traditionally, in the production of Scandinavian ropy milks, low temperature incubation between l3°C to l8°C is preferred, because incubation at temperatures higher than 27°C to 30°C resulted in considerable reduction or loss of desirable high viscosity and mucoidness (Bottazzi, V., Other Fermented Dairy Products. p. 328, 345-346. In: G. Reed (ed.), Biotechnology-Vol.5, Food and Feed Production with Microorganisms. Verlag Chemie, Weinheim, Federal Republic of Germany (l983); and Macura, D., et al. J. Dairy Sci. 67:735-744(l984)). It had been suggested by early prior art that the mucoidness might protect the lactic Streptocococcus against bacteriophage; however, this was shown to be wrong. Sozzi et al, Milchwissenschaft 33, 349-352 (l978).
The association of several metabolic functions in lactic streptococci with plasmid DNA is now well recognized (McKay, L. L., J. Microbiol. 49:259-274 (l983)). On the basis of the observed instability of ropy characteristic in lactic streptococci, Macura and Townsley (Macura, D., et al., J. Dairy Sci. 67:735-744 (l984)) and McKay suggested that plasmid DNA may be involved in the expression of mucoid phenotype (Muc⁺).
A problem in the prior art is to be able to produce phage resistant strains of Streptococcus which are members of the N group. It would be highly desirable to be able to impart phage resistance to stra ins of Streptococcus which are phage sensitive since these bacteria are very important in commercial fermentations for producing fermented milk products. McKay et al, Applied Environmental Microbiology, 47; 68-74 (l984) describes limited phage resistance which is plasmid associated. Klaenhammer, J., Advances in Applied Microbiology 30, l-29 (l984) at page 22 discusses plasmid encoded phage resistance. Phage resistance has not been associated with a l8.5 Mdal plasmid in Streptococcus cremoris encoding for mucoidness. Further, Streptococcus cremoris NRRL-B-l5995 was obtained as a single colony isolated from a phage resistant strain but is a slow acid producer and thus is not a useful strain for milk fermentations.

Objects

It is therefore an object of the present invention to provide a method for imparting phage resistance to Streptococcus of the N group which are phage sensitive. Further it is an object of the present invention to provide novel phage resistant bacteria derived from phage sensitive strains. These and other objects will become increasingly apparent by reference to the following description and the drawings.

In the Drawings

Figure l is a drawing of an agarose gel electrophoresis of plasmid DNA from Streptococcus cremoris MS and its cured derivatives showing an l8.5 Mdal plasmid which encodes for mucoid substance production. These are as follows: (A) Parent strain MS (8) MS0l (C) MS02 (D) MS03 (E) MS04 (F) MS05 and (G) Reference plasmid DNA for molecular sizing from Eschericia coli V5l7.
Figure 2 is a drawing of an agarose gel electrophoresis of plasmid DNA from (A) S. lactis ML-3/2.2 (B) S. lactis ML-3/2.20l (C) S. cremoris MS (D) S. cremoris MS04 (E) S cremoris MS040l (F) S lactis ML-3/2.202 and (G) Reference plasmid DNA for molecular sizing from E. coli V5l7.
Figure 3 is a drawing of an agarose gel electrophoresis of plasmid DNA from (A) S. lactis ML-3/2.202 (B) Malty S. lactis 4/4.20l (C) S. lactis subsp. diacetylactis SLA3.25 (D) S. lactis subsp. diacetylactis SLA3.25ul and (E) Reference plasmid DNA for molecular sizing from E. coli V5l7.

General Description

The present invention relates to a method for imparting phage resistance to Streptococcus bacteria which comprises: providing a phage sensitive bacteria of the genus Streptococcus group N which is lysed by a homologous phage; and introducing a transferred plasmid into the phage sensitive bacteria to thereby produce a phage resistant bacteria which is resistant to the homologous phage, wherein the transferred plasmid contains DNA derived from a parental plasmid which encodes for a mucoid substance around the outside of Streptococcus cremoris (MS) NRRL-B-l5995.
Further the present invention relates to a phage resistant bacteria of the species Streptococcus lactis or Streptococcus lactis subspecies diacetylactis in substantially pure form derived from a phage sensitive bacteria and containing plasmid DNA derived from a parental plasmid which encodes for a mucoid substance from Streptococcus cremoris (MS) NRRL-B-l5995, wherein the phage resistant bacteria is resistant to a homologous phage, and to heat cured phage resistant derivatives of the phage resistant bacteria wit h the plasmid integrated into bacteria so as to be unidentifiable, generally into the chromosomes of the bacteria.
The present invention also relates to a phage resistant bacteria of the species Streptococcus lactis or Streptococcus lactis subspecies diacetylactis which were derived from phage sensitive parent cells by conjugal transfer of a plasmid which encodes for a mucoid substance in Streptoccocus cremoris (MS) NRRL-B-l5995. Further, the present invention relates to heat cured Muc⁻ derivatives of the phage resistant transconjugants lacking the l8.5 Mdal plasmid which still retain resistance to homologous phages.
The bacterial cells can be prepared for use as a concentrate having a pH between about 4 and 8 and containing at least about l × l0⁷ cells per gram up to about l0¹⁵ cells per gram, usually between about l × l⁰⁹ and l0¹² cells per gram. The concentrates can be frozen with or without a freezing stabilizing agent such as monosodium glutamate, malt extract, non-fat dry milk, alkali metal glycerophosphate, glutamic acid, cystine, glycerol, or dextran or the like and then thawed for use or the concentrates can be lyophilized or dried by other means to a powder as is well known to those skilled in the art. The bacterial cells are generally used in a range between about l0⁵ to l0⁹ cells per ml of milk to be fermented, depending upon the product to be produced. All of this is very well known to those skilled in the art. U.S. Patent No. 3,420,742 describes various preservation methods.
U.S. Patent No. 4,382,097 to one of the inventors herein describes mixed cultures including mucoid substance producing (Muc⁺) strains. The phage resistant, mucoid substance producing strains of the present invention can be used in the preparation of these mixed cultures with good results.

Specific Description

The following Example shows the involvement of plasmid DNA (Muc plasmid) in the expression of the Muc⁺ phenotype in Streptococcus cremoris MS. Additionally, the Example shows the conjugal transfer of Muc-plasmid from a ropy Streptococcus cremoris to a non-mucoid (Muc⁻) Streptococcus lactis and from the resultant mucoid Streptococcus lactis transconjugant to a malty variant of Streptococcus lactis (formerly Streptococcus lactis var. maltigenes) and a strain of Streptococcus lactis subsp. diacetylactis and the expression of Muc⁺ phenotype in all the transconjugants. The resulting transconjugant Streptococcus lactis and Streptococcus lactis subsp. diacetylactis strains are phage resistant. The phage susceptibility of the malty Streptococcus lactis transconjugant was not determined because of the unavailability of a lytic phage for the parent strain. It is believed to be phage resistant.
Streptococcus cremoris MS (NRRL-B-l5995) when grown in milk at 24°C, ferments lactose (Lac⁺) and produces a mucoid (ropy) coagulum (Muc⁺). Streptcococcus cremoris MS was isolated by the inventor from milk products it is not available from any other source. By incubating Streptococcus cremoris MS at 38°C, several non-mucoid (Muc⁻) isolates were obtained. Comparison of plasmid profiles of mucoid and non-mucoid isolates showed that a l8.5 Mdalton plasmid (pSRQ2202) was involved in the expression of mucoid phenotype. Additionally, the curing experiments revealed that in Streptococcus cremoris MS, a 75.8 Mdalton plasmid (pSRQ220l) was associated with the ability to ferment lactose. Derivatives lacking pSRQ220l did not ferment lactose (Lac⁻). In mating experiments using Streptococcus cremoris MS as donor, pSRQ220l was conjugatively transferred to Lac⁻ Streptococcus lactis ML-3/2.2. The transconjugant Streptococcus lactis ML-3/2.20l was Lac⁺, which confirmed that pSRQ220l coded for lactose utilization.
By indirect selection techniques using genetic markers for lactose utilization or phage resistance, pSRQ2202 was first conjugatively transferred from a Lac⁻, Muc⁺ derivative of Streptococcus cremoris MS (Strain MS04) to Lac⁺, Muc⁻ Streptococcus lactis ML-3/2.20l. The resultant transconjugant Streptococcus lactis ML-3/2.202 was Lac⁺ and Muc⁺. Subsequently, pSRQ2202 was co-mobilized with pSRQ220l in mating experiments, from Streptococcus lactis ML-3/2.202 (NRRL-B-l5996) (donor) to a plasmid-free, Lac⁻, Muc⁻ malty Streptococcus lactis 4/4.2, and a Lac⁻, Muc⁻ Streptococcus lactis subsp. diacetylactis SLA3.25. The respective transconjugants were Lac⁺ and Muc⁺ confirming that pSRQ220l and pSRQ2202 encoded for Lac⁺ and Muc⁺ phenotypes respectively. With the transfer of pSRQ2202, the transconjugants Streptocococcus lactis ML-3/2.202 (NRRL-B-l5996) and Streptococcus lactis subsp. diacetylactis SLA3.250l (NRRL-B-l5994) and l8-l6.0l (NRRL-B-l5997 not only acquired Muc⁺ phenotype but also resistance to phages, which were lytic to respective parent strains, namely Streptococcus lactis ML-3/2.20l and Streptococcus lactis subsp. diacetylactis SLA3.25 and l8-l6. The strains marked with an NRRL number were deposited with the Northern Regional Research Laboratory, Peoria, Illinois and are freely available to those who request them by name and number.

Example 1

MATERIALS AND METHODS

Cultures and Phages: Bacterial strains used in this study are listed in Table l.
Phage c2 lyses Streptococcus lactis C₂ and Streptococcus lactis ML-3. Phage 643 is lytic for Streptococcus lactis ML-3 and lytic phage (designated phage l8-l6) plaques on Streptococcus lactis subsp. diacetylactis l8-l6 and SLA 3.25.
Culture Media and Propagation: Cultures that fermented lactose (Lac⁺) were routinely propagated in sterile l0% reconstituted non-fat dry milk (NFM) at 24°C for l4-l6 hours. Strains that were lactose negative (Lac⁻) were grown in sterile NFM fortified with 0.5% glucose and 0.2% yeast extract (FNFM).
Stock cultures grown in NFM or FNFM containing l0% sterile glycerol as cryoprotectant were dispensed into cryogenic vials and stored in liquid nitrogen. For routine use, additional vials of cultures were stored in a freezer held at -60°C.
In curing experiments to eliminate Muc⁺ phenotype either NFM or FNFM was initially used. Later, BMG broth (Gonzalez, C. F., et al., Appl. Environ. Microbiol. 46:8l-89 (l983)) was used, and for the selection and purification of Lac⁻ colonies, BML agar (BMLA) described by Gonzalez and Kunka was used. Milk-indicator agar (MIA) was used for the selection and purification of phenotypes expressing variations of Lac and Muc (Lac ^+/-Muc ^+/-) in curing and mating experi ments. MIA was made up in two separate parts which after sterilization and tempering at 60°C were mixed together. The first part consisted of distilled water at half the final volume of the medium (final volume l liter), containing the required amount of non-fat milk solids to give a final concentration of 5%; the second part consisted of the remaining half of distilled water containing the required amount of agar to obtain l.5% in the final medium, and l0.0 ml of 0.8% aqueous solution of bromocresol purple.
Donor and recipient cultures for mating were grown to logarithmic phase (6-8 hours at 24°C) in either whey broth (Lac⁺ strains) or whey-glucose broth (Lac⁻ strains). Whey broth (WB) was made in two parts, sterilized separately and after cooling, mixed together. To make up l liter of WB, 70.0g of sweet whey powder (Pallio Dairy Products Corp., Campbell, N.Y.) was dissolved in 500 ml distilled water and centrifuged to remove undissolved residue. To the clear supernatant, l9.0 g sodium-beta-glycero phosphate (Sigma Chemical Co., St. Louis, MO) was added, mixed well and sterilized at l2l°C for l5 minutes. The second portion of the medium consisted of 5.0 g yeast extract, l0.0 g tryptone, 5.0 g gelatin, 0.5 g sodium acetate, 0.5 g MgSO₄.7H₂O and 0.2 g CaCl₂.2H₂O dissolved in 500 ml distilled water. After sterilization at l2l°C for l5 minutes the two parts were mixed together when cool (at 50°C). Whey-glucose broth (WBG) was made up by including 5.0 g of glucose in the formulation for WB. Matings were performed on 5% milk-glucose agar (MGA) plates as described by McKay et al (McKay, L. L., et al., Appl. Environ. Microbiol. 40:84-9l (l980)).
Curing: For temperature curing, S. cremoris strains were incubated between 38°C and 39°C; Streptococcus lactis and Streptococcus lactis subsp. diacetylactis strains were incubated between 4l°C and 42°C. For eliminating Lac⁺ phenotype, cultures inoculated at the rate of 0.05% in BMG were incubated overnight at elevated temperatures and plated at suitable dilutions onto BMLA plates. Presumptive white Lac⁻ colonies were confirmed for inability to ferment lactose and purified by single colony isolation on BMLA. For eliminating Muc⁺ phenotype, cultures inoculated at 0.05% in NFM/FNFM or BMG were incubated at elevated temperatures overnight and suitable dilution plated on MIA plates. Individual colonies were picked either into NFM (Lac⁺) or FNFM (Lac⁻) and incubated at 24°C until coagulation or thickening of milk occurred. The cultures were then tested for mucoidness with l.0 ml graduated serological pipets. Resistance to easy flow and formation of stringiness (long ropy strands) during free fall from pipet tip were used as test criteria to establish mucoidness.
Mating: In each mating experiment, two donor: recipient ratios (i.e. l:2 and l:4) were used. Mating mixtures and respective donor and recipient controls spread on MGA plates were incubated overnight at 24°C in a Gas Pak anaerobic jar (BBL Microbiology Systems, Cockeysville, MD). Cells from the surface of MGA plates were harvested with sterile Basal broth (BM, Gonzalez and Kunka set forth previously), using l.0 ml BM per plate. Washings from each set of plates containing one experimental variable (i.e., plates with donor cells or recipient cells or mating mixture l:2 or mating mixture l:4) were pooled together, centrifuged and resuspended in l.0 ml BM. The entire suspension was then plated onto five BMLA plates (0.2 ml per plate) containing appropriate concentration of selective drugs. Streptomycin (Sm) was added to obtain a final concentration of l000 micrograms per ml; fusidic acid (Fus) to a final concentration of 2u micrograms per ml and Rifampin (Rif) to a final concentration of 300 micrograms per ml. Plates were incubated at 24°C for 72 hours and examined. All Lac⁺ colonies were transferred to NF M to test for mucoidness. Mucoid isolates were purified by single colony isolation on MIA and subjected to confirmatory tests.
In mating experiments where specific lytic phage was used as selective agent, pelleted cells harvested from MGA plates were resuspended in 2.0 ml of high titer phage lysate (>l0⁷ PFU/ml). One-tenth ml of 0.2 M CaCl₂ was added to each tube, mixed gently and allowed to stand for 15 to 20 minutes at room temperature for phage adsorption. The contents of each tube were then spread on lO BMLA plates (0.2 ml per plate).
For donor input counts, dilutions were plated on BMLA (Lac⁺) or BMGA (Lac⁻). Platings for counts were made immediately before mating mixtures and controls were spread on MGA. Transfer frequencies were calculated as the number of mucoid colonies per donor colony-forming-unit (CFU). Mating experiments were repeated at least once, and in some cases twice. To exciude transduction as a possible mode of genetic transfer, in parallel mating mixtures, donor culture aliquots were replaced with an equal amount of cell-free filtrates (Millipore filter, 0.45 micron-pore size, Millipore Corp. Bedford, MA) of the donor. To exclude transformation, DNAse at a final concentration of l00 micrograms per ml was added to the mating mixture prior to plating and to the MGA used. As a negative control to show that live cells are needed for the observed genetic transfer, in parallel mating mixtures, donor culture aliquots were replaced with heat-killed (boiling water-bath for l0 minutes) donor culture portions.
Confirmatory Tests: Confirmatory testing was done for arginine hydrolysis (Niven, C. F., Jr., et al., J. Bacteriol. 43:65l-660 (l942)), diacetyl-acetoin production from citrate in milk (King, N., Dairy Industries l3:800 (l948)), and susceptibility to specific phages by spot test on seeded agar-overlay plates. Where necessary resistance or sensitivity to additional drug markers not selected for in the mating protocols was determined.
Cell lysis and Electrophoresis: For rapid screening of strains for plasmids, the method described by Anderson and McKay (Anderson, D. G., et al. Appl. Environ. Microbiol. 46:549-552 (l983)) was used. Procedures described by Gonzalez and Kunka (set forth previously) were used for routine examination of plasmid DNA, and for preparing purified plasmid DNA using cesium chloride gradients.

EXPERIMENTAL PROCEDURES

Curing of Muc⁺ Phenotvpe:

Initial experiments with Streotococcus cremoris MS showed that mucoidness (or ropiness) in this strain could be easily eliminated by incubating inoculated NFM tubes at 38°C for l4 to l6 hours. In three separate trials, an average of 30% of colonies isolated from MA plates spread with Streptococcus cremoris MS which had been incubated at elevated temperatures in NFM, were non-mucoid. Twenty mucoid and 2u non-mucoid milk coagluating (in 18 to 24 hours at 24°C) isolates were purified and examined for plasmid DNA content by agarose gel electrophoresis. All mucoid types with one exception, namely strain MS03, showed plasmid profiles similar to that of the wild type mucoid Streptococcus cremoris MS (Figure l). In Streptococcus cremoris MS03 the 35.8 Mdalton plasmid was absent. All the non-mucoid isolates with one exception showed plasmid profiles similar to Streptococcus cremoris MS0l (Figure l). In these strains, two plasmids, namely the 35.8 Mdalton and l8.5 Mdalton plasmids were missing. In the single non-mucoid isolate, Streptococcus cremoris MS02, only the l8.5 Mdalton plasmid was absent (Figure l). By examining the plasmid profiles of Streptococcus cremoris strains MS, MS0l, MS02 and MS03, only the l8.5 Mdalton plasmid ( Q2202) could possibly be associated with Muc⁺ phenotype because strain MS03 retained the Muc⁺ phenotype even with the loss of 35.8 Mdalton plasmid. On the other hand, strain MS02, which possessed the larger 35.8 Mdalton plasmid became non-mucoid with the elimination of pSRQ2202.
Curing of Lac⁺ Phenotype: It is now well established that lactose fermenting ability in lactic streptococci is plasmid borne (McKay, L. L., Regulation of lactose metabolism in dairy streptococci, p. l53-l82. In: R. Davis (ed.), Developments in Food Microbiology -l. Applied Science Publishers Ltd., Essex, England (l982)). To determine if Lac-plasmid in Muc⁺ Streptococcus cremoris MS03 could be cured without the loss of mucoid or ropy characteristic, the culture was incubated at 38°C for l6 hours and plated on BMLA. Of a total of 80 colonies appearing on the BMLA plates, l5% were Lac⁻. All the Lac⁻ colonies were transferred into the FNFM, incubated at 24°C until thickening or coagulation occurred. The cultures were checked for mucoidness at that stage. With the exception of two isolates, all others were non-mucoid. Streptococcus cremoris MS04 represents the Lac⁻ Muc⁺ phenotype (Figure l). Analysis of the plasmid profiles of Lac-cured derivatives suggested that the loss of lactose fermenting ability was associated with the elimination of 75.8 Mdalton plasmid (pSRQ220l). The Lac⁻ Muc⁻ phenotype is represented by Streptococcus cremoris MS05.
Development of Strategies for Transfer of Muc-plasmid: Alongside curing experiments, mucoid and non-mucoid strains were extensively examined for differences in carbohydrate fermentation patterns, resistance to different levels of NaCl, bile salts, ethanol, nisin and antioxidants like butylated hydroxyanisole (BHA) and butylated hydroxytoluene (BHT).
No differences were observed that could be used to rapidly screen for mucoid types in the presence of non-mucoid types on agar plates. Further experiments in agar media containing various dyes and varous levels and combinations of sugar, milk components and minerals (Mg⁺⁺, Ca⁺⁺, Mn⁺⁺, Fe⁺⁺⁺) did not yield a good differential medium for visual distinction of mucoid from non-mucoid colonies. It was then decided to attempt indirect selection for possible mucoid types in genetic experiments. Because conjugative transfer of plasmids in lactic streptococci is well documented (Gasson, M. J., J. Microbiol. 49:275-282 (l983)), we decided to examine if Lac-plasmid in mucoid strains could be used as a metabolic marker to detect co-mobilization of Muc-plasmid in mating experiments.
Conjugative Transfer of Lac-plasmid: Several attempts to transfer Lac⁺ phenotype from Streptococcus cremoris MS (Lac⁺ Muc⁺ Sm ^s; superscript s denotes "sensitive" throughout the text) to Streptococcus cremoris MS05.2 (Lac⁻ Muc⁻ Fus ^r Sm ^r) and to the plasmid-free Streptococcus lactis SLA l.l (Lac⁻ Muc⁻ Sm ^r) proved unsuccessful. In a later mating experiment, Lac⁺ phenotype was transferred from Streptococcus cremoris MS to Streptococcus lactis ML-3/2.2 (Lac⁻ Muc⁻ Sm ^r Fus ^r). The transconjugant Streptococcus lactis ML-3/2.20l was Lac⁺ but Muc⁻ and had acquired a 75.8 Mdalton plasmid from Streptococcus cremoris MS. Additionally, the transconjugant was positive for arginine hydrolysis test and susceptible to phages c2 and 643. The expression of Lac⁺ phenotype by Streptococcus lactis ML-3/2.20l as a result of the acquisition of 75.8 Mdalton plasmid, pSRQ220l from the donor, and the curing data obta ined with Streptococcus cremoris MS and its derivatives indicated that pSRQ 220l coded for lactose utilization in Streptococcus cremoris MS. Results from three independent mating experiments reconfirmed the association of Lac⁺ phenotype in transconjugants to the acquisition of pSRQ220l from Streptococcus cremoris MS. Some of the transconjugants obtained in these matings exhibited clumping in broth cultures similar to the phenomenon previously reported by Walsh and McKay (Walsh, P. M., et al., J. Bacteriol. 146:937-944 (l98l)). Parallel control experiments conducted with one of the matings to exclude transformation and transduction and negative control experiment using heat-killed donor Streptococcus cremoris MS cells showed that the transfer of Lac-plasmid was by conjugation.
Conjugative Transfer of Lac- and Muc-plasmids:
A clumping Lac⁺ transconjugant Streptococcus lactis ML-3/2.20l was used as donor in a mating with Streptococcus cremoris MS04 to determine if the Lac-plasmid pSRQ220l observed in the transconjugant Streptococcus lactis ML-3/2.20l could be transferred back to Streptococcus cremoris MS04, a Lac⁻ Muc⁺ derivative of the wild type. The selective agent for this mating was phage c2 (at a titer of l x l0⁴ PFU/ml) which lyses the donor, Streptococcus lactis ML-3/2.20l. The phage does not infect the recipient Streptococcus cremoris MS04. Because of the low titer of phage used for selecting against donor cells, control donor plates showed several Lac⁺ survivor colonies per plate. Plates containing the mating mixtures, however, had at least twice as many Lac⁺ colonies per plate as the control donor plates. All acid-producing colonies from BMLA plates containing the mating mixtures were transferred into NFM and after incubation, checked for mucoidness. Randomly chosen 200 non-mucoid isolates were purified and checked for resistance to streptomycin (l000 micrograms per ml) on BMLA. All were resistant; additionaly, they gave positive tests for liberation of NH₃ from arginine indicating that they were donor types. Six isolates that were mucoid were purified on MIA and checked for streptomycin sensitivity or resistance, and for arginine hydrolysis. Two of the mucoid isolates were negative for arginine hydrolysis and the other four were positive. Repurified isolates of the six mucoid cultures showed the same arginine hydrolysis characteristics as in the first testing. The two mucoid isolates that failed to liberate NH₃ from arginine were also sensitive to streptomycin (l000 micrograms per ml) and fusidic acid (20 micrograms per ml) indicating that they were recipient type Lac⁺ transconjugants. The remaining four mucoid cultures that were positive for arginine hydrolysis were resistant to the same levels of streptomycin and fusidic acid as the donor, Streptococcus lactis ML-3/2.20l. All six mucoid isolates were resistant to phages c2 and 643. Agarose gel electrophoresis of plasmid DNA from the six isolates showed that two types of transconjugants were obtained. The two arginine-negative, streptomycin sensitive Lac⁺ Muc⁺ isolates were recipient Streptococcus cremoris MS04 types that had acquired the Lac-plasmid (and Lac⁺ phenotype) from the donor (exemplified by Streptococcus cremoris MS040l, Figure 2). The four arginine-positive, Lac⁺ Muc⁺ isolates were donor Streptococcus lactis ML-3/2.20l-type transconjugants that had acquired the l8.5 Mdalton Muc-plasmid from Streptococcus cremoris MS04 (exemplified by Streptococcus lactis ML-3/2.202, Figure 2). Phage resistance of Muc ⁺ transconjugants of donor Streptococcus lactis ML-3/2.20l type suggested that the acquisition of l8.8 Mdalton plasmid pSRQ2202 and Muc⁺ phenotype conferred virus resistance to the ML-3/2.202 transconjugant, although the parent strain ML-3/2.20l (Lac⁺ Muc⁻) was susceptible to the same phage. Based on these initial observations, the mating was repeated with Streptococcus cremoris MS04 as the donor and using a high titer c2 phage lysate (3.0 x l0⁹ PFU per ml) to select effectively against Lac⁺, phage-sensitive recipient Streptococcus lactis ML-3/2.20l cells.
Colony-free recipient control plates were obtained. Because the use of high titer phage lysate provided effective selection against non-mucoid, Lac⁺, phage-sensitive recipient cells, Lac⁺ colonies appearing on mating plates probably were transconjugants of donor Streptococcus cremoris MS04 type that had acquired the Lac-plasmid from Streptococcus lactis ML-3/2.20l or were Muc⁺, phage-resistant transconjugants of ML-3/2.20l type that had acquired the Muc-plasmid from Streptococcus cremoris MS04. Based on that premise, all Lac⁺ colonies from BMLA plates containing the mating mixtures were transferred into NFM and tested for mucoidness. All mucoid isolates were purified and subjected to confirmatory arginine hydrolysis test. With the use of high titer phage lysate transfer of Muc⁺ phenotype was observed at a frequency of 3.0 × l0⁻⁴ Parallel control experiments conducted to exclude transformation and transduction and negative control experiment using heat-killed donor Streptococcus cremoris MS04 ceils showed that the mode of transfer of Muc-plasmid from donor to recipient was conjugative.
Incubation of the Lac⁺ Muc⁺ transconjugant Streptococcus lactis ML-3/2.202 at 40 to 42°C allowed the selection of all possible combinations of Lac ^+/- Muc ^+/- phenotypes. Agarose gel electrophoretic profiles of plasmid DNA from such derivatives confirmed that Lac⁺ phenotype was expressed when the 75.8 Mdalton pSRQ220l plasmid was present and in the absence of that plasmid the bacteria were Lac⁻. Similarly, the presence and absence of l8.5 Mdalton pSRQ2202 plasmid was directly associated with the expression of Muc⁺ and Muc⁻ phenotypes, respectively.
To further characterize the l8.5 Mdalton Muc-plasmid it was necessary to obtain the specific covalently closed circular DNA in a purified form. To obtain pSRQ2202 isolated by itself in parentalor transconjugant strains, it would be necessary to cure several other resident pasmids in those strains. Aternatively, the Muc-pasmid coud be transferred to a plasmid-free strain singly or in association with another plasmid - for example, Lac-plasmid - which then could be cured out leaving only the Muc-plasmid. Streptococcus lactis LM0230 and its plasmid-free derivatives have been successfully used as recipients for facilitating such transfer of desired plasmid singly or in association with another plasmid which could be cured out subsequently leaving only the desired plasmid in the recipient (McKay, L. L. et al., Appl. Environ. Microbiol. 47:68-74 (l984)). Accordingly, a mating experiment was conducted with Streptococcus cremoris 040l (Lac⁺ Muc⁺ transconjugant) as the donor and Streptococcus lactis SLAl.l (plasmid-free Lac⁻ Muc⁻ Sm ^r) as the recipient. Selection for transconjugants was based on Lac⁺ phenotype and streptomycin resistance. Streptococcus cremoris 040l was chosen as the donor because it would allow analysis for an unselected marker (arginine hydrolysis) in presumptive transconjugant colonies. The MS040l X SLA l.l mating resulted in the transfer of Lac⁺ phenotype at a frequency of 2.0 × l0⁻³; however, there was no transfer of Muc⁺ phenotype. Rapid screening of 200 Lac⁺ purified isolates from the mating plates that were arginine positive for plasmid DNA profiles revealed that in the majority of the isolates, in addition to the Lac-plasmid some of the resident cryptic plasmids in Streptococcus cremoris MS040l were co-transferred. Attempts to co-transfer Muc-plasmid with the Lac-plasmid from Streptococcus lactis ML-3/2.202 to plasmid-free Streptococcus lactis SLA i.8 (Lac⁻ Muc⁻ Rif ^r) proved unsuccessful.
Co-transfer of Muc-plasmid with Lac-plasmid: As an alternative to Streptococcus lactis LM0230 derivatives, a plasmid-free, malty Streptococcus lactis 4/4.2 (obtained by curing two resident plasmids from the wild type malty strain Streptococcus lactis 4) was used as recipient in a mating with Streptococcus lactis ML-3/2.202. Malty Streptococcus lactis 4/4.2 had Rif ^r and Fus ^r markers. The wild type malty Streptococcus lactis 4 gave a strong Lac⁺ reaction on BMLA and coagulated milk in 10 to l2 hours at 24°C. The plasmid-free, malty derivative Streptococcus lactis 4/4.2 exhibited a weak Lac⁺ reaction BMLA when incubated longer than 48 hours at 24°C and failed to coagulate milk after 48 hours at 24°C. The indicator response in BMLA to threshold level of lactose utilization by Streptococcus lactis 4/4.2 was eclipsed by the addition of 0.5% disodium beta-gycerophosphate to BMLA. Plating of the mating mixtures on the buffered medium containing the selective antibiotic allowed the selection of Lac⁺ transconjugants. Lactose positive purified isolates screened for sensitivity to streptomycin were tested for Muc⁺ phenotype. Presumptive Lac⁺ Muc⁺ transconjugants were confirmed by testing for production of malty odor in milk containing sodium salt of 4 methyl-2 oxopentanoic acid (Langeveld, L.P.H. Neth. Milk Dairy J. 29:l35 (l975)). Mobilization of other cryptic plasmids in addition to Muc-plasmid was observed with the transfer of Lac-plasmid from Streptococcus lactis ML-3/2.202 to Streptotoccus lactis 4/4.2 (Figure 3). In repeat experiments with suitable controls, transfer of Lac- and Muc-plasmids was established to be conjugative. It was of interest to determine if Muc-plasmid could be co-transferred with Lac-plasmid to the third member of the lactic Streptococcus group; i.e., Streptococcus lactis subsp. diacetylactis. Mating experiments using Streotococcus lactis ML-3/2.202 as the donor and Streptococcus lactis subsp. diacetylactis SLA3.25 as the recipient were conducted. These matings yielded co-transfer of Lac- and Muc-plasmids to Streptococcus lactis subsp. diacetylactis; and, in some cases, transfer of only the Lac-plasmid was observed. The recipient Streptococus lactis subsp. diacetylactis SLA3.25 and the Lac⁺ transconjugants were sensitive to phage l8-l6. The Lac⁺ Muc⁺ transconjugant Streptococcus lactis subsp. diacetylactis SLA3.250l, however, was resistant to phage l8-l6. By incubating the Lac⁺ Muc⁺ transconjugant SLA3.250l at 4l to 42C, all possible combinations of Lac and Muc phenotypes (Lac ^+/- Muc ^+/-) were obtained. As observed with Streptococcus lactis ML-3/2.202, the elimination of pSRQ220l and pSRQ2202 SLA3.250l correlated with the inability of the derivatives to express Lac⁺ and Muc⁺ phenotypes respectively. In repeat experiments with suitable controls, co-transfer of Lac- and Muc-plasmids from Streptococcus lactis ML-3/2.202 to Streptococcus lactis subsp. diacetylactis SLA3.25 was confirmed to be conjugative.
Table 2 summarizes inter-species transfer frequencies for pSRQ220l and pSRQ2202 singly and for the co-transfer of pSRQ2202 with the lac-plasmid from Streptococcus lactis ML-3/2.202 to malty Streptococcus lactis 4/4.2 and Streptococcus lactis subsp. diacetylactis SLA3.25.
The results of Example l presented clearly demonstrate that mucoid phenotype in the lactic streptococci examined is encoded on plasmid DNA. The association of mucoid phenotype with plasmid DNA in the wild type S. cremoris MS was initially demonstrated by curing experiments. In these experiments, the presence or absence of l8.5 Mdalton plasmid correlated with mucoid and non-mucoid phenotypes respectively. The actual confirmation that mucoid phenotype is encoded on plasmid DNA was obtained in mating experiments, where the conjugative transfer of l8.5 Mdalton plasmid from mucoid S. cremoris MS04 to non-mucoid S. lactis ML-3/2.20l enabled the transconjugant containing the l8.5 Mdalton pSRQ2202 plasmid to express the mucoid phenotype. Additionally, the elimination of pSRQ2202 from the mucoid transconjugant S. lactis ML-3/2.202 resulted in a non-mucoid phenotype. Subsequently, pSRQ2202 was conjugatively transferred to S. lactis subsp. diacetyactis SLA 3.25 S. lactis 4/4.2. The phenotypic expresson of pSRQ2202 in the respective transconjugants ( S. lactis 4/4.20l and SLA 3.250l) indicate that in general, mucoid phenotype in lactic streptococci is linked to plasmid DNA.
The ease with which the Muc-plasmid was eliminated by incubating between 38C and 42C was in keeping with the earlier observation that to retain the desired mucoid characteristic in Scandinavian ropy milks, low temperature incubation between l3C to l8C is favored; incubation at temperatures higher than 27°C to 30°C resulted in considerable reduction or loss of desirable high viscosity and mucoidness.
In addition to the Muc-plasmid, transfer of Lac-plasmid was achieved. The Lac-plasmid from the wild-type mucoid S. cremoris MS was first transferred to Streptococcus lactis ML-3/2.2 and subsequenty the same plasmid was retransferred from Streptococcus lactis ML-3/2.20l to the Lac⁻, Muc⁺ derivative of the wild type mucoid Streptocccus cremoris ( S. cremoris MS04). Further, the plasmid was transferred from the Lac⁺ Muc⁺ transconjugant S. lactis ML-3/2.202 to the malty S. lactis 4/4.2 and S. lactis subsp. diacetylactis SLA 3.25. In the latter two matings, the Lac-plasmid also mobilized the Muc-plasmid and other cryptic plasmids. In all these transfers, the Lac⁺ phenotype was expressed in the respective transconjugants and the elimination of 75.8 Mdalton plasmid from the respective transconjugants rendered them Lac⁻.
Although the transfer of Muc-plasmid was detected in these mating experiments using indirect selection procedures, namely, scoring for Lac⁺ phenotype and/or phage-resistance, direct selec tion procedure through the use of other differential medium to distinguish between mucoid and non-mucoid colony types is possibie. All of this is well known to those skiiled in the art.
A significant finding in Example l was the association of phage resistance and mucoidness. With the transfer of Muc-plasmid to a non-mucoid, phage-sensitive recipient, the resultant mucoid transconjugant became resistant to the phage. This held true with Streptococcus lactis and Streptococcus lactis subsp. diacetylactis. The association of phage-resistance with mucoid phenotype in transconjugants offers another mechanism whereby phage-resistant derivatives for starter cutures can be made. Additionally, the seiection procedure for the distinction of transconjugants through the use of high titer lytic phage lysates provides a means for avoiding drug markers for selection. This is especialiy significant in deriving desired strains for food and feed fermentations. It was also found that even if the Muc⁺ plasmid was cured from the transconjugant by high temperature incubation, the resulting strains were phage resistant although they had lost the ability to produce the mucoid substance. It appeared that at least a portion of the plasmid integrated with chromosomal material or with another part of the cell. This is a desirable method for fixing phage resistance into the bacteria cells.

Example 2

A well known non-mucoid commercial strain Streptococcus cremoris TR was rendered mucoid by transferring Muc⁺ phenotype from Lac⁻ derivative of mucoid Streptococcus lactis transconjugant ML-3/2.202. Phage tr which is lytic for S. cremoris TR was used to select against non-mucoid, phage-sensitive, Lac⁺ recipient cells. Only Lac⁺ survivor colonies from mating plates were picked into milk and tested for mucoidness. Mucoid cultures were purified on MIA and reexamined for mucoidness in milk, phage resistance, arginine hydrolysis, and subjected to plasmid analysis.
Mucoid transconjugant TR0l was resistant to phage tr, did not hydrolyze arginine and was Lac⁺.
Transconjugant TR0l was cured to Muc⁻ phenotype by high temperature incubation. The non-mucoid derivative retained resistance to phage tr.
The mucoid transconjugant and its non-mucoid cured derivative have no antibiotic markers and are suitable for food fermentations. If the bacteria are to be used in foods, selection is made for strains which are antibiotic sensitive.

Example 3

Muc-plasmid from Lac-cured derivative of Streptococcus lactis ML-3/2.202 was transferred to non-mucoid, Lac⁺, phage-sensitive Streptococcus lactis subsp. diacetylactis l8-l6 using phage l8-l6 as selecting agent. Only Lac⁺ colonies were selected to examine for mucoidness. Mucoid cultures were purified and reexamined for mucoidness, phage-resistance and subjected to confirmatory King's test.
Transconjugant Streptococcus lactis subsp. diacetylactis l8-l6.0l does not have any antibiotic markers, and was Lac⁺, positive for diacetyl-acetoin production in milk, resistant to phage l8-l6, and mucoid. Comparison of plasmid profiles of parent Streptococcus lactis subsp. diacetylactis l8-l6, transconjugant Streptococcus lactis subsp. diacertylactis l8-l6.0l and donor Lac-cured Streptococcus lactis ML-3/2.202 showed that the mucoid transconjugant had acquired a 25 Mdalton plasmid, which coded for Muc⁺ phenotype. Apparently the l8.5 Mdal plasmid acquired some additional DNA through a recombinational event. The association of Muc⁺ phenotype with 25 Mda l plasmid was confirmed by curing studies.
The application of the mucoid Streptococcus lactis subsp. diacetylactis l8-l6.0l transconjugant in Cottage cheese cream dressing was examined. Dry Cottage cheese curd is mixed with sufficient cream dressing to obtain stipulated milk fat content to meet legal specifications. The cream dressing may be cultured to develop diacetyl flavor and may contain hydrocolloid stabilizers (e.g., agar, carageenan, gums, and the like) to increase the viscosity of cream dressing so that it will adhere to cured surface rather than settling down to the bottom of the container. The use of flavor cultures in Cottage cheese dressing to develop diacetyl flavor and to increase shelf-life is well known in the industry. The use of the phage resistant, mucoid producing Streptococcus lactis subspecies diacetylactis in Cottage cheese creaming mixtures or other milk containing products to replace stabilizers is unknown. There are many naturally mucoid producing strains.
The use of mucoid Streptococcus lactis subsp. diacetylactis l8-l6.0l in Cottage cheese dressing was examined for the following:

1. If half-and-half cream (l8% milk fat) cultured with Streptococcus lactis subsp. diacetylactis l8-l6.0l (l% inoculum, l6 hr. at 74°F) is comparable in viscosity to commercial, non-cultured, stabilized Cottage cheese dressing.
2. If half-and-half cream cultured with strain l8-l6.0l has better flavor characteristics than uncultured commercial dressing.
3. If half-and-half cream cutured with strain l8-l6.0l when used as dressing at a dry curd: dressing ratio of 64:36 exhibits the same level of adherence to curd as commercial, stabilized, uncultured dressing used at the same ratio.
4. If cheese curd dressed with cultured half-and-half cream using strain l8-l6.0l had better flavor (diacetyl) and keeping quality than cheese dressed with uncultured, stabilized, commercial dressing (stored at 40°F-45°F).

Dry Cottage cheese curd and uncultured, commercial cream dressing containing stabilizer mixture consisting of guar gum, carageenan, and locust bean gum and fungal inhibitor potassium sorbate were obtained from a local supplier. Half-and-half cream containing no additives was purchased from a local supermarket.
Dry Cottage cheese curd was washed in lightly chlorinated ice water and drained to remove excess water. A portion of half-and-half cream was steamed (in freely flowing steam in a chamber) for 30 minutes, cooled to 74°F and cultured with Streptococcus lactis subsp. diacetylactis l8-l6.0l (l% inoculum from a milk culture) for l6 hours at 74°F. After incubation the cultured half-and-half cream was chilled in an ice-bath. A psychrotrophic culture of Pseudomonas fragi PFO, isolated from spoiled Cottage cheese was grown overnight at 76°F in Trypticase soy broth. The broth culture was diluted in sterile dilution buffer to obtain l x l0⁶ to l x l0⁷ cells per milliliter. The experimental variables were set up in the following manner:

1. 330g dry curd + l70g commercial dressing.
2. 330g dry curd + l70g half-and-half cultured with Streptococcus lactis subsp. diacetylactis l8-l6.0l.
3. 330g dry curd + l70g half-and-half.
4. 330g dry curd + l70g commercial dressing + Pseudomonas fragi PFO cells to give about l x l0⁴ cells per gram of curd.
5. 330g dry curd + l70g half-and-half + Pseudomonas fragi PFO at about l x l0⁴ cells per gram of curd.
6. 330g dry curd + l70g half-and-half cultured with strain l8-l6.0l plus strain PFO added at about l x l0⁴ cell per gram of curd. Before preparing the Cottage cheese samples, uncultured half-and-half, cultured half-and-half and commercial dressing were tested for viscosity using Zahn cup #2.

Dressed curd at 500g portions prepared according to the experimental design were made up in duplicates and distributed into duplicate plastic cartons. Cross contamination was avoided in all the operations. All the ingredients were kept cold in an ice-bath during the various operations. Packaged cartons were transferred to a walk-in cooler that was controlled at 40°F. At weekly intervals, one set of cartons representing the experimental variables were examined visually for spoilage and by smelling for development of diacetyl flavor or the lack or loss of developed diacetyl flavor, and for off-flavors. At the end of four-week period the duplicate, unopened set of cartons representing the experimental variables were checked and the results were recorded.

Results:

Viscosity Measurements:
Uncultured half-and-half = less than l00 centipoises
Stabilized commercial dressing = l50 centipoises
Half-and-half containing no stabilizer and cultured with Strain l8-l6.0l = l80
centipoises
The results showed that culturing with Strain l8-l6.0l protected the cheese from psychrotrophic spoilage (variable 5 versus variable 6). Culturing with strain l8-l6.0l also enhanced flavor (variable l and 4 versus variables 2 and 6). Comparable protection against spoilage by P. fragi PFO between variables 4 and 6 may be attributed to the presence of potassium sorbate in the commercial dressing. The cultured half-and-half did not contain any sorbate and the entire inhibitory activity was due to S. lactis subsp. diacetylactis.
The unopened cartons examined after 4 weeks showed similar results to the set examined at weekly intervals.
The cultured half-and-half had as good a viscosity and curd adhering property as stabilized, uncultured commercial creaming mixture. The use of mucoid S. lactis subsp. diacetylactis l8-l6.0l eliminates the addition of stabilizers in the creaming mixture. Additionally, it provides good diacetyl flavor and increased shelf-life.
The strains can also be used in Cottage cheese containing fruits or fruit puree. The mucoidness can form a barrier around the curd and keep it separate from any fruit used in the Cottage cheese and prevent moisture loss from the curd due to osmotic pressure differential between the fruit and cheese curd.
Recently, a patent (Vedamuthu, E. R., et al., U.S. Patent No. 4,382,097 (l983)), was issued for the application of ropy strains of Streptococcus lactis and/or Streptococcus cremoris in combination with non-mucoid cultures in specific proportions, for the production of non-ropy cultured dairy products possessing good viscosity and heavy body. Commercial concentrated cultures containing ropy and non-ropy lactic streptococci for the production of Cultured Buttermilk and Sour Cream are currenty available in the United States. Such combination cultures help to maintain a thick, heavy body in fermented dairy products without the use of hydrocolloid stabilizers or fortification with milk solids. The phage resistant and mucoid substance producing strains of the present invention can be used in these mixed cultures.
It will be appreciated that the l8.5 Mdal plasmid can be introduced into the sensitive strain by transformation or transduction or by bacterial conjugation. These techniques are well known to those skilled in the art.

Claims

1. A method for impaarting phage resistance to Streptococcus bacteria which comprises:

(a) provi ding a phage sensitive bacteria of the genus Streptococcus group N which is lysed by a homologous phage; and

(b) introducing a transferred plasmid into the phage-sensitive bacteria, to thereby produce a phage resistant bacteria which is resistant to the homologous phage, wherein the transferred plasmid contains DNA derived from a parental plasmid which encodes for production of a mucoid substance around the outside of Streptococcus cremoris (MS) NRRL-B-l5995.

2. The method of Claim l wherein the parental plasmid is carried in Streptococcus cremoris NRRL-B-l5995 (MS).

3. The method of Claim l wherein the transferred plasmid is introduced into the phage sensitive bacteria by mating with a strain containing the parental plasmid.

4. The method of Claim l wherein the sensitive bacteria is selected from the group consisting of Streptococcus lactis, Streptococcus lactis subsp. diacetylactis, and Streptococcus cremoris.

5. The method of Claim l wherein a Lac plasmid coding for lactose fermentation is co-transferred into the sensitive bacteria with the transferred plasmid which encodes for the mucoid substance.

6. The method of Claim 5 wherein the transferred plasmids are carried in a Streptococcus lactis which is mated with the sensitive bacteria.

7. The method of Claim l wherein the parental plasmid is carried in Streptococcus cremoris NRRL-B-l5995 (MS) and wherein the sensitive bacteria is selected from a Streptococcus lactis and wherein the parental plasmid is transferred by mating.

8. The method of Claim l wherein the transferred plasmid is carried in a Streptococcus lactis which is mated with the sensitive bacteria, wherein the sensitive bacteria is selected from a Streptococcus lactis subsp. diacetylactis and wherein the transferred plasmid is transferred by mating.

9. The method of Claim l wherein the sensitive bacteria is a Streptoccocus lactis subspecies diacetylactis.

10. The method of Claim l wherein the resistant bacteria is selected from Streptococcus lactis NRRL-B-l5996, Streptococcus lactis subsp. diacetylactis NRRL-B-l5994 (SLA 3.250l) and Streptococcus lactis subsp. diacetylactis NRRL-8-l5997.

11. The method of Claim l including the additional step of heat curing the bacteria so that the ability to produce the mucoid resistance is lost and the phage resistance is retained.

12. The method of Claim l wherein the transferred plasmid is used as a marker for selection of transconjugant bacteria from phage sensitive bacteria.

13. A phage resistant bacteria of the species Streptococcus lactis or Streptococcus lactis subspecies diacetylactis in substantially pure form is derived from a phage sensitive bacteria and containing plasmid DNA derived from a parental plasmid which encodes for a mucoid substance from Streptococcus cremoris (MS) NRRL-B-l5995, wherein the phage resistant bacteria is resistant to a homologous phage, and to heat cured phage resistant derivatives of the phage resistant bacteria with the plasmid integrated into bacteria so as to be unidentifiable.

14. The bacteria of Claim l3 selected from Streptococcus lactis NRRL-B-l5996 and Streptococcus lactis subsp. diacetylactis NRRL-B-l5994 and NRRL-B-l599 7.

15. A method of increasing the thickness of milk products without addition of stabilizers which comprises:

(a) providing in a milk containing product a derived bacteria of the species Streptococcus lactis or Streptococcus lactis subspecies diacetylactis containing transferred plasmid containing DNA derived from a parental plasmid which encodes for a mucoid substance from Streptococcus cremoris (MS) NRRL-B-l5995; and

(b) incubating the milk containing product with the phage resistant bacteria to develop a mucoid substance and to increase the thickness of the milk containing product.

16. The bacteria of Claim 15 selected from Streptococcus lactis NRRL-B-l5996 and Streptococcus lactis subsp. diacetylactis NRRL-B-l5994 and NRRL-B-l5997.

17. The method of Claim l5 wherein the milk containing product is a Cottage cheese creaming mixture.

18. The method of Claim l5 wherein the derived bacteria are phage resistant.